Separate Head Research Articles

Research of anomaly detection based on dynamic anomaly detection enhancement framework

Abstract Anomaly detection plays a crucial role in various fields, from industrial defect inspection to geological detection. However, traditional approaches often struggle with insufficient discriminability and an inability to generalize to unseen anomalies. These limitations stem from the practical difficulty in gathering a comprehensive set of anomalies and the tendency to overlook anomalous instances in favor of normal samples. To address these challenges, we propose a novel Dynamic Anomaly Detection Enhancement Framework, integrating three key innovations: (1) SaliencyAug: An adaptive saliency-guided augmentation method that generates realistic pseudo-samples to enhance learning of rare anomalies, improving model generalization. (2) DynAB: A dynamic attention block that achieves effective multi-level feature fusion while minimizing redundant information, enhancing detection accuracy. (3) DualOM: A dual-head optimization module which employs separate heads for normal and anomalous sample learning, creating more explicit and discriminative decision boundaries. Extensive experiments across multiple real-world datasets demonstrate our framework's superior performance in detecting a wide range of anomalies, demonstrating 2.4% improvement over state-of-the-art methods.

Comparison Between InterCriteria and Correlation Analyses over sEMG Data from Arm Movements in the Horizontal Plane

InterCriteria analysis (ICrA) and two kinds of correlation analyses, Pearson (PCA) and Spearman (SCA), were applied to surface electromyography (sEMG) signals obtained from human arm movements in the horizontal plane. Ten healthy participants performed ten movements, eight of which were cyclic. Each cyclic movement (CM) consisted of flexion and extension phases with equal duration (10 s, 6 s, 2 s, and 1 s) and two 5 s rest poses between them. The CMs were performed with and without an added load of 0.5 kg on the wrists of the participants. The sEMG signals from six different muscles or separate muscle heads (m. deltoideus pars clavicularis, m. deltoideus pars spinata, m. brachialis, m. anconeus, m. biceps brachii, and m. triceps brachii long head) were recorded and used to compare the results of the ICrA, PCA, and SCA. All three methods found identical consonance pairs for the flexion and extension CM phases. Additionally, PCA detected two more consonance pairs in the extension phases. In this investigation, ICrA, PCA, and SCA were proven to be reliable tools when applied separately or in combination for sEMG data. These three methods are appropriate for researching arm movements in the horizontal plane and experimental protocol revision.

Possible Representations of Aker in Heb-sed Scenes Adorning Temple Entryways and their Cosmic Significance

Summary This paper examines the unique representations of the separated heads on the ground line on the columns and gateways in the temple entryways associated with the Sed-festival. An analysis of the scenes shows that here are mutual iconographical and religious characteristics between this motif and the chthonic deity Aker. The focal point of the research is to study the known representations of these heads and present an additional hypothesis of attributing them to Aker.

25 Years of Türkiye’s Media Initiatives in Africa

Since around 2005, Türkiye has focused on establishing close relations with African countries, spanning politics, economics, culture, and media. This study aims to briefly touch upon the transformation of press and media in Africa and examine Türkiye’s initiatives in the African media landscape, particularly since 2005. This original work aims to serve as a guide for both media professionals and decision-makers in Türkiye’s foreign policy. Additionally, it seeks to ensure that the country’s media relations with African countries are based on a solid foundation. Due to the broad scope of the subject, Türkiye’s media initiatives have been analyzed under separate headings. Moreover, Türkiye’s steps in media relations with African countries, especially since 2005, have been investigated. To gather information on the topic, news articles, official statements, agreements signed by Türkiye with African countries in the media field, and decisions made during this process have been examined. Upon examination of the country’s media relations with African countries, it is observed that various initiatives have been taken, particularly since 2008. These relations in the media field have seen an upward trend through the 2020s. However, there is a need for further steps to be taken, particularly in areas such as the employment of African media professionals and collaboration in the media field, as well as producing more content related to Africa.

Feature decomposition-based gaze estimation with auxiliary head pose regression

Recognition and understanding of facial images or eye images are critical for eye tracking. Recent studies have shown that the simultaneous use of facial and eye images can effectively lower gaze errors. However, these methods typically consider facial and eye images as two unrelated inputs, without taking into account their distinct representational abilities at the feature level. Additionally, implicitly learned head pose from highly coupled facial features would make the trained model less interpretable and prone to the gaze-head overfitting problem. To address these issues, we propose a method that aims to enhance task-relevant features while suppressing other noises by leveraging feature decomposition. We disentangle eye-related features from the facial image via a projection module and further make them distinctive with an attention-based head pose regression task, which could enhance the representation of gaze-related features and make the model less susceptible to task-irrelevant features. After that, the mutually separated eye features and head pose are recombined to achieve more accurate gaze estimation. Experimental results demonstrate that our method achieves state-of-the-art performance, with an estimation error of 3.90° on the MPIIGaze dataset and 5.15° error on the EyeDiap dataset, respectively.

Changes of bone and articular cartilage in broilers with femoral head necrosis

Femoral head necrosis (FHN) in broilers is a common leg disorder in intensive poultry farming, giving rise to poor animal health and welfare. Abnormal mechanical stress in the hip joint is a risk factor for FHN, and articular cartilage is attracting increasing attention as a cushion and lubrication structure for the joint. In the present study, broilers aged 3 to 4 wk with FHN were divided into femoral head separation (FHS) and femoral head separation with growth plate lacerations (FHSL) groups, with normal broilers as control. The features of the hip joint, bone, and cartilage were assessed in FHN progression using devices including computed tomography (CT), atomic force microscope (AFM), and transmission electron microscopy (TEM). Broilers with FHN demonstrated decreased bone mechanical properties, narrow joint space, and thickened femoral head stellate structures. Notably, abnormal cartilage morphology was observed in FHN-affected broilers, characterized by increased cartilage thickness and rough cartilage surfaces. In addition, as FHN developed, cartilage surface friction and friction coefficient dramatically increased, while cartilage modulus and stiffness decreased. The ultramicro-damage occurred in chondrocytes and the extracellular matrix (ECM) of cartilage. Cell disintegration, abnormal mitochondrial accumulation, and oxidative stress damage were observed in chondrocytes. A notable decline in cartilage collagen content was observed in ECM during the initial stages of FHN, accompanied by a pronounced reduction in collagen fiber diameter and proteoglycan content as FHN progressed. Furthermore, the noticeable loosening of the collagen fiber structure and the appearance of type I collagen were noted in cartilage. In conclusion, there was a progressive decrease in bone quality and multifaceted damage of cartilage in the femoral head, which was closely linked to the severity of FHN in broilers.

EUS-guided hydrogel injection to separate pancreatic head carcinoma from duodenum for enhanced radiotherapy: Multi-site feasibility study.

Background and study aims The proximity of a pancreas head tumor to the duodenum often limits delivery of an ablative dose of radiation therapy. This study evaluated the feasibility and safety of using an injectable polyethylene glycol (PEG) hydrogel between the head of the pancreas and duodenum. Patients and methods In a multi-site feasibility cohort study of patients with localized pancreatic cancer, PEG hydrogel was injected under endoscopic ultrasound guidance to temporarily position the duodenum away from the pancreas. Procedure characteristics were recorded, including hydrogel volume and space created. Patients were monitored for adverse events (AEs) and radiotherapy toxicity. Results In all six intent-to-treat patients (four with borderline resectable, two with locally advanced disease), the ability to place and visualize PEG hydrogel and create space between the duodenum and the head of the pancreas was successful. There were no procedure-related AEs resulting in radiotherapy delay. There were no device-related AEs and no reports of pancreatitis. Conclusions PEG hydrogel was successfully placed, created space between the duodenum and the head of the pancreas, and was not associated with major toxicity. Enhancing radiotherapy for pancreatic cancer by using PEG hydrogel to create peri-duodenal space could have beneficial implications for treatment and warrants more exploration.

HMMED: A Multimodal Model with Separate Head and Payload Processing for Malicious Encrypted Traffic Detection

HMMED: A Multimodal Model with Separate Head and Payload Processing for Malicious Encrypted Traffic Detection

РОЗРОБКА ЕФЕКТИВНОГО ТА ТОЧНОГО МЕТОДУ ВІДСТЕЖЕННЯ ОБʼЄКТІВ НА ВІДЕО НА МОБІЛЬНИХ ПРИСТРОЯХ

One of the top research topics in computer vision area is visual object tracking. Main goal is to obtain the target object’s location in the first video frame of a given video sequence. The recent innovations of deep neural networks, specifically Siamese networks has significant impact on visual object tracking. In spite of high accuracy and high results in academic benchmarks, there are drawbacks in current state-of-the-art approaches in particular compute-intensive and large memory footprint that cannot satisfy the performance requirements of real-world applications. The aim of this paper is to design a new lightweight framework for resource-efficient and accurate visual object tracking. To add, a new tracker of efficiency benchmark and protocol were introduced. Efficiency is defined in terms of both energy consumption and execution speed on edge devices. New dual template representation for object model adaptation was developed. The first template, static, fixes the original visible appearance and thus prevents deviation and, as a result, failures caused by adaptation. The other is dynamic; the state reflects the current conditions of assembly and the appearance of its object. Unlike STARK, which incorporates additional timing information by introducing a separate estimation prediction head, we introduce parameter-free module similarity as a template update rule optimized from the latest network. We show that the learned convex combination of two patterns is effective for tracking on multiple tests. A lightweight tracker was proposed, which includes functions, dual representation of patterns and pixel-by-pixel merged blocks in its compact network. The resulting FEAR-XS tracker runs at 205 FPS on the iPhone 11, which is 4.2 times faster than LightTrack and 26.6 times faster than Ocean, with high accuracy on many tests – no state-of-the-art tracker is more accurate and faster than any FEAR tracker. In addition, the algorithm is highly energy efficient.

TR-DETR: Task-Reciprocal Transformer for Joint Moment Retrieval and Highlight Detection

Video moment retrieval (MR) and highlight detection (HD) based on natural language queries are two highly related tasks, which aim to obtain relevant moments within videos and highlight scores of each video clip. Recently, several methods have been devoted to building DETR-based networks to solve both MR and HD jointly. These methods simply add two separate task heads after multi-modal feature extraction and feature interaction, achieving good performance. Nevertheless, these approaches underutilize the reciprocal relationship between two tasks. In this paper, we propose a task-reciprocal transformer based on DETR (TR-DETR) that focuses on exploring the inherent reciprocity between MR and HD. Specifically, a local-global multi-modal alignment module is first built to align features from diverse modalities into a shared latent space. Subsequently, a visual feature refinement is designed to eliminate query-irrelevant information from visual features for modal interaction. Finally, a task cooperation module is constructed to refine the retrieval pipeline and the highlight score prediction process by utilizing the reciprocity between MR and HD. Comprehensive experiments on QVHighlights, Charades-STA and TVSum datasets demonstrate that TR-DETR outperforms existing state-of-the-art methods. Codes are available at https://github.com/mingyao1120/TR-DETR.

Changes of bone remodeling, cartilage damage and apoptosis-related pathways in broilers with femoral head necrosis 1

Femoral head necrosis (FHN) is a common leg disorder in the poultry industry often leads to significant cartilage damage. The mechanism behind abnormal apoptosis in FHN broilers, leading to cartilage damage, remains unclear; although endoplasmic reticulum stress (ERS) has been found to play a role in glucocorticoid-induced FHN broilers. In this study, we collected samples from broilers with femoral head separation (FHS) and femoral head separation accompanied with growth plate lacerations (FHSL) in a broiler farm. The aim was to investigate the potential association between the severity of FHN, bone remodeling and cartilage damage. Additionally, primary chondrocytes were treated with methylprednisolone (MP) to construct an in vitro FHN model, followed by inhibition or activation of ERS or hypoxia inducible factor-1α (HIF-1α) to further investigate the mechanism of apoptosis in cartilage. The results suggested that cartilage appeared to be the appropriate tissue to investigate the potential mechanisms of FHN, as the degree of cartilage damage was found to be closely related to the severity of the disease. Bone quality was only affected in FHSL broilers, although factors related to bone metabolism were significantly altered among FHN-affected broilers. In addition, cartilage in FHN-affected broilers exhibited high levels of apoptosis and upregulated expression of ERS-related and HIF-1α, which was consistent with both in vivo and in vitro findings after MP treatment. The results were further supported by treatment with HIF-1α or ERS inhibition or activation. In conclusion, bone remodeling and cartilage homeostasis were affected in FHN broilers, but only cartilage damage was significantly exacerbated with FHN development. Moreover, activation of ERS or HIF-1α resulted in apoptosis in cartilage, thus exhibiting a significant correlation with FHN severity.

Multiagent Continual Coordination via Progressive Task Contextualization.

Cooperative multiagent reinforcement learning (MARL) has attracted significant attention and has the potential for many real-world applications. Previous arts mainly focus on facilitating the coordination ability from different aspects (e.g., nonstationarity and credit assignment) in single-task or multitask scenarios, ignoring the stream of tasks that appear in a continual manner. This ignorance makes the continual coordination an unexplored territory, neither in problem formulation nor efficient algorithms designed. Toward tackling the mentioned issue, this article proposes an approach, multiagent continual coordination via progressive task contextualization (MACPro). The key point lies in obtaining a factorized policy, using shared feature extraction layers but separated independent task heads, each specializing in a specific class of tasks. The task heads can be progressively expanded based on the learned task contextualization. Moreover, to cater to the popular centralized training with decentralized execution (CTDE) paradigm in MARL, each agent learns to predict and adopt the most relevant policy head based on local information in a decentralized manner. We show in multiple multiagent benchmarks that existing continual learning methods fail, while MACPro is able to achieve close-to-optimal performance. More results also disclose the effectiveness of MACPro from multiple aspects, such as high generalization ability.

Positive effects of health behaviors acquired during the COVID-19 pandemic process on the prevention of other infectious diseases.

It was aimed to evaluate the positive effects of health behaviors (general hygiene, wearing face masks, physical distancing, and travel restrictions) acquired during the coronavirus disease 2019 (COVID-19) pandemic on the prevention of other infectious diseases in Ankara Province, Türkiye. This study was designed retrospectively. Among the notifiable group A infectious diseases, acute intestinal infections (AIIs) with International Classification of Diseases, Tenth Revision diagnosis codes A09 (diarrhea and gastroenteritis presumed to be of infectious origin), R11 (nausea and vomiting), and K52 (other noninfectious gastroenteritis and colitis), as well as influenza, tuberculosis, measles, varicella, malaria, and meningococcal meningitis were included in the scope of this study.The data of the selected infectious diseases in Ankara Province for the last 2 years before the pandemic (January 2018-December 2019) and for the 2-year period of the pandemic (January 2020-December 2021) were analyzed after checking the data. The number of cases were presented as frequencies, the 1-sample chi-squared test was used in the statistical analysis and the statistical significance level (α) was taken as 0.05. The findings for each disease/disease group were discussed under separate headings. Comparing the prepandemic period (2018-2019) with the pandemic period (2020-2021), the decreases in the number of cases of selected infectious diseases, except influenza, were statistically significant. Undoubtedly, the experience gained from the pandemic struggle will guide us in shaping our future lives. From this point forward, we should be aware that living in crowded environments and as a highly mobile population, that unhygienic habits are unfavorable for the spread of all infectious diseases, and we should take care to continuously apply the precautions for healthy living in our new lifestyle.

Evaluation of Patients aged over 65 years who applied to Ankara Training and Research Hospital Family Medicine Outpatient Clinics: Comparison Before and After Covid-19 Pandemic

Aim: Taking into account the importance and priority of the geriatric age group, the aim of our study was to examine the reasons for and the number of visits to primary health care institutions, to compare them before and after the Covid-19 pandemic, and to draw attention to the importance of geriatric sciences in family medicine specialty training in the light of this evaluation.
 
 Materials and Methods: Patients who applied to the Ankara Training and Research Hospital (EAH) Family Medicine Outpatient Clinics between 01.03.2020-31.12.2022 constitute the population of the study. Patients were retrospectively evaluated before and after the Covid-19 pandemic according to their admission rates and diagnoses. The outpatient clinics they applied to were analyzed. Musculoskeletal system problems were analyzed under a separate heading according to diagnoses.
 Results: Out of 47014 diagnoses above the age of 65 during the pandemic period, I10 - Essential (Primary) Hypertension was the most common diagnosis, except for general medical examinations and screening examinations. In the post-Covid-19 period, 45928 diagnoses were made in 14398 patients over the age of 65. After general medical and laboratory examinations, the most common diagnose was I10 Essential Hypertension in both period. Musculoskeletal complaint during the pandemic period was the second most common with the diagnosis of M79.1 Myalgia in both period. Musculoskeletal complaints especially myalgia was more common in the pandemic period than the post-covid period (p

A Pilot Study of Simulation-Free Hippocampal-Avoidance Whole Brain Radiotherapy Using Diagnostic MR-Based and Online Adaptive Planning

We aimed to demonstrate the clinical feasibility and safety of a simulation-free hippocampal avoidance whole brain radiotherapy (HA-WBRT) workflow in a Phase I clinical trial (NCT05096286). Feasibility was defined as successful completion of the simulation-free HA-WBRT workflow through treatment delivery in at last 70% of treated patients. Ten candidates for HA-WBRT were enrolled for treatment on a ring gantry CT-guided Linac with online adaptive capabilities. Structures were contoured on the diagnostic brain MRI, which was then registered to a separate head computed tomography (CT) of similar head shape, obtained from an atlas-based database. A HA-WBRT "pre-plan" was generated using the atlas-based CT (AB-CT) and the NRG-CC001 constraints. At first fraction, the AB-CT was used as the primary dataset and deformed to the patient's cone-beam CT (CBCT) for dose calculation. The brain, ventricles, and brainstem contours were matched through rigid translation and rotation to the corresponding anatomy on the CBCT to aid in alignment, given the differences in rotational head positioning from diagnostic MRI to CBCT setup. Lastly, the lens, optic nerves, and brain contours were manually edited based on CBCT visualization. Plans were then optimized, and the adaptive plan was chosen for treatment if the plan met all objectives. Workflow tasks were timed. In addition, conventional plans using patients' sim CTs were created for each patient for the purpose of prospective dosimetric comparison. The dosimetric parameters were compared for each patient between the delivered sim-free plan and the conventional sim CT plan using the sign test via statistical software, with p<.05 indicating significance. Median time from approved sim order to first fraction was 4 days (range: 2-7); median time in room (door-to-door) was 49 minutes (range: 35-70). All patients successfully completed all ten fractions and 90% of the simulation-free radiation plans met all NRG-CC001 constraints. For one patient, the sim-free plan at fraction one failed the planning target volume (PTV) coverage objective (coverage of 89%); this was deemed acceptable for delivery by the treating radiation oncologist. An offline replan was then performed to meet NRG-CC001 constraints and used for the subsequent nine fractions. There was no clinically meaningful difference in dosimetric constraints between the sim-free plan (calculated on AB-CT) and conventional CT sim plan. Statistically, the sim-free plans provided improved PTV coverage to higher doses compared to the conventional plans (Table). At a median follow-up of 43 days (range: 9 -280), the intracranial progression-free survival rate was 90%. Simulation-free HA-WBRT is feasible, results in plans that are dosimetrically comparable to conventional CT sim workflows and succeeds in decreasing time to initiation of HA-WBRT by at least 50%. Further studies with a larger cohort are warranted to optimize the workflow.

Learning radiologists’ annotation styles with multi-annotator labeling for improved neural network performance

BACKGROUND: One of the common problems in labeling medical images is inter-observer variability. The same image can be labeled differently by doctors. The main reasons are the human factor, differences in experience and qualifications, different radiology schools, poor image quality, and unclear instructions. The influence of some factors can be reduced by proper organization of the annotation; however, the opinion of doctors frequently differs.&#x0D; AIM: The study aimed to test whether a neural network with an additional module can learn the style and labeling features of different radiologists and whether such modeling can improve the final metrics of object detection on radiological images.&#x0D; METHODS: For training artificial intelligence systems in radiology, cross-labeling, i.e., annotation of the same image by several doctors, is frequently used. The easiest way is to use labeling from each doctor as an independent example when training the model. Some methods use different rules or algorithms to combine annotation before training. Finally, Guan et al. use separate classification heads to model the labeling style of different doctors. Unfortunately, this method is not suitable for more complex tasks, such as detecting objects on an image. For this analysis, a machine learning model designed to detect objects of different classes on mammographic scans was used. This model is a neural network based on Deformable DETR architecture. A dataset consisting of 7,756 mammographic breast scans and 12,543 unique annotations from 19 doctors was used to train the neural network. For validation and testing, a dataset consisting of 700 and 300 Bi-Rads-labeled scans, respectively, was taken. In all data sets, the proportion of images with pathology was in the 15%20% range. A unique index was assigned to each of the 19 doctors, and a special module at each iteration of the neural network training found a vector corresponding to this index. The vector was expanded to the size of the feature map of each level of the feature pyramid, and then attached by separate channels to the maps. Thus, the encoder and the decoder of the detector had access to the information about which doctor labeled the scan. The vectors were updated using the back-propagation method. Three methods were chosen for comparison:&#x0D; &#x0D; Basic model: Combining labels by different doctors using the voting method.&#x0D; New stylistic module: For predictions on the test dataset, a single doctors index was taken, which showed the best metrics on the validation dataset.&#x0D; New stylistic module: The indexes of the five doctors with the best metrics on the validation dataset were used for predictions on the test dataset. Weighted Boxes Fusion was chosen to combine the predictions.&#x0D; &#x0D; The area under the receiver operating characteristic curve (ROC-AUC) was used as the primary metric on the test dataset (Bi-Rads 3, 4, and 5 categories were referred to pathology). The sum of maximum probabilities of detected malignant objects (malignant masses and calcinates) by cranio-caudal and medio-lateral oblique projections was assumed as the probability of malignancy for each method.&#x0D; RESULTS: The following ROC-AUC metrics were obtained for the three methods: 0.82, 0.87, and 0.89.&#x0D; CONCLUSIONS: The information about the labeling doctor allows the neural network to learn and model the labeling style of different doctors more effectively. In addition, this method may obtain an estimate of the uncertainty of the networks prediction. The use of embedding from different doctors, leading to different predictions, may mean that this data is difficult for an artificial intelligence system to process.

The unilateral muscular slip of the two-bellied Serratus Posterior Inferior muscle: a case report and a review of literature.

The two-bellied serratus posterior inferior (SPI) muscle with a muscular slip is a rare variation of the back region, which can cause significant discomfort to patients. Patients typically present with symptoms of chronic pain syndrome, radiating back pain, myofascial pain, or lower back pain. This report describes a case of a female cadaver with a two-headed SPI muscle and a right muscular slip, along with a literature review. A case of an unusual back muscle was observed in a female cadaver during advanced cadaver dissection of the back region. The SPI muscle was found deep to the latissimus dorsi muscle and superficial to the erector spinae and thoracolumbar fascia. Its oblique arrangement and insertion onto the 8th-11th costae were consistent with its known anatomy, but the presence of two separate fibrotendinous heads and an uncommon variance between the erector spinae and latissimus dorsi muscles was observed. The SPI muscle fibers were found to be attached to the 8th costa on the right side and displayed two heads on both sides. In our study, the presence of muscular and tendinous digitations towards the twelfth rib was not detected, which aligns with the characteristics described for types D and E. However, we observed a separation between the digitations. Therefore, it is appropriate to classify our findings as type E according to the established categorization. Simultaneously, an anomalous muscular slip, which did not conform to any of the aforementioned classifications, was identified as extending towards the 8th rib. The underlying cause of unilateral oblique muscular fiber extension is thought to result from aberrant muscle migration during embryonic development or alterations in tendon attachment sites. Differential diagnosis of unidentified lower back pain should consider the various types and alterations of the SPI muscle.

Reason-able embeddings: Learning concept embeddings with a transferable neural reasoner

We present a novel approach for learning embeddings of ALC knowledge base concepts. The embeddings reflect the semantics of the concepts in such a way that it is possible to compute an embedding of a complex concept from the embeddings of its parts by using appropriate neural constructors. Embeddings for different knowledge bases are vectors in a shared vector space, shaped in such a way that approximate subsumption checking for arbitrarily complex concepts can be done by the same neural network, called a reasoner head, for all the knowledge bases. To underline this unique property of enabling reasoning directly on embeddings, we call them reason-able embeddings. We report the results of experimental evaluation showing that the difference in reasoning performance between training a separate reasoner head for each ontology and using a shared reasoner head, is negligible.

Are modular pedicle screws associated with a high complication rate following posterior spinal fixation?

Modular pedicle screws have a separate head that can be intraoperatively assembled to the inserted shank. The aim of this study was to report associated intra- and post-operative complications and reoperation rates of posterior spinal fixations with modular pedicle screws at a single center. A retrospective, institutional chart review was performed on 285 patients who underwent posterior thoracolumbar spinal fusion with modular pedicle screw fixation between January 1, 2017, and December 31, 2019. The primary outcome was failure of the modular screw component. Other measures recorded were length of follow-up, other complications, and need for additional procedures. There were 1,872 modular pedicle screws (average 6.6 per case). There were no (0.0%) screw head dissociations at the rod screw junction. There was 20.8% overall complication rate (59/285) with 25 reoperations: 6 due to non-union and rod breakage, 5 for screw loosening, 7 for adjacent segment disease, 1 for acute postoperative radiculopathy, 1 for epidural hematoma, 2 for deep surgical-site infections, and 3 for superficial surgical-site infections. Other complications included superficial wound dehiscence [8], dural tears [6], non-unions not requiring reoperation [2], lumbar radiculopathies [3], and perioperative medical complications [5]. This study demonstrates that modular pedicle screw fixation has reoperation rates similar to those previously reported for standard pedicle screws. There was no failure at the screw-head junction, and no increases in other complications. Modular pedicle screws present an excellent option to allow surgeons to place pedicle screws without the risk of extra complications.

Application Effect Analysis for Helmet Detection Algorithm based on YOLOV5

Safety helmets can provide better safety protection for the human head in industrial production, which is one of the important protective tools to protect the lives of workers. In the on-site environment, checking whether construction workers are wearing safety helmets mainly depends on manual inspection, which is time-consuming and labor-intensive. In order to solve the problem of hard hat detection and recognition in the complex environment of the construction site, in this paper, we propose a YOLOV5 hard hat detection method. Specifically, a pre-trained object model is used to locate the area of the worker and the face detection module is used to mark the face. Then, the helmet recognition module is used to extract the sub area of the worker's head. Finally, the extracted image is classified into two categories to judge the situation of the worker wearing the helmet. Focusing only on workers can reduce training time and improve the efficiency of the model, so the training label only uses the label of people. The model refines and extracts the head, inputs the separated head area into the subsequent network for subsequent analysis, and then uses the binary classification method to judge whether the helmet is worn. We compare the results of different models such as YOLOV5s, YOLOV5m and YOLOV5l, where YOLOV5l can achieve a best mAP of 91.9%. We also analyze the effect of light to the performance, where the detection effect will be affected by low visibility through the comparative detection of samples in the day and at night.